eduzhai > Applied Sciences > Engineering >

Performance characterization of a novel deep learning-based MR image reconstruction pipeline

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 18 pages

Abstract: A novel deep learning-based magnetic resonance imaging reconstructionpipeline was designed to address fundamental image quality limitations ofconventional reconstruction to provide high-resolution, low-noise MR images.This pipeline s unique aims were to convert truncation artifact into improvedimage sharpness while jointly denoising images to improve image quality. Thisnew approach, now commercially available at AIR Recon DL (GE Healthcare,Waukesha, WI), includes a deep convolutional neural network (CNN) to aid in thereconstruction of raw data, ultimately producing clean, sharp images. Here wedescribe key features of this pipeline and its CNN, characterize itsperformance in digital reference objects, phantoms, and in-vivo, and presentsample images and protocol optimization strategies that leverage image qualityimprovement for reduced scan time. This new deep learning-based reconstructionpipeline represents a powerful new tool to increase the diagnostic andoperational performance of an MRI scanner.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×